All eukaryotes use three essential DNA-dependent RNA polymerases to decode genetic information stored in DNA, namely RNA Polymerases I, II and III. Remarkably, plants have two more RNA polymerases, abbreviated as Pol IV and Pol V, that evolved from Pol II to specialize in the synthesis of RNAs that guide gene silencing, an important process in all living organisms for controlling retrotransposons, viruses or genes important for development. Pols IV and V are key to a complicated RNA-directed DNA methylation (RdDM) pathway. The pathway is initiated by Pol IV, acting in partnership and physical association with a RNA- dependent RNA polymerase, RDR2 to transcribe DNA into short double-stranded RNAs (dsRNAs). These dsRNAs are then trimmed, from either end, by the DICER endonuclease, DCL3, yielding 24 nt short interfering RNAs (siRNAs). The siRNAs, loaded into ARGONAUTE 4 (AGO4), guide the siRNA-AGO complexes to sites of Pol V transcription, where they bind to Pol V transcripts as well as to the C-terminal domain of the Pol V largest subunit. The DNA methyltransferase, DRM2 (the ortholog of human DNMT3) is recruited, methylating cytosines within the Pol V-transcribed DNA. Resulting heterochromatin formation is refractive to transcription by Pols I, II or III. However, Pols IV and V are not repressed in this chromatin environment. Instead, helper proteins that recognize cytosine methylation or repressive histone modifications are thought to recruit Pols IV and V, taking the place of transcription factors and dispensing with conventional promoter elements. There is much that we do not understand about the RdDM process. What unwinds DNA for Pols IV and V to gain access to template strands? How are Pol IV and RDR2 activities coupled for dsRNA synthesis? Where do Pol V transcripts begin and end, and how many siRNA-AGO complexes can bind them ? What do the presumed Pol IV and Pol V helper proteins actually do? By devising new biochemical assays combined with genetic, genomic and structural studies, our goal is to answer these questions in mechanistic detail. In eukaryotes as diverse as humans, flies, worms and fission yeast, noncoding RNAs guide chromatin modifications important for centromere function, transposon silencing, X-chromosome inactivation or imprinting of paternal or maternal alleles. Of special relevance to our studies is the piRNA pathway that directs the silencing of transposons in the human germline, thereby serving the same purpose, and using the same DNA methylation machinery, as the RdDM pathway of plants. Controlled DNA methylation is critical, such that aberrant DNA methylation and chromatin modification is implicated in the pathology of Rett, ICF, Prader-Willi, Beckwith-Wiedemann and Fragile X syndromes, and in most forms of cancer. By understanding the biogenesis and targeting mechanisms of noncoding RNAs in DNA methylation and gene silencing, our studies will contribute new understanding of fundamental processes important for human development and disease.